This project made progress in several areas within the past year, including: (1) providing further evidence that a specific part of the medial temporal lobe (MTL) functions in both perception and memory, contrary to the prevailing, memory-only view; (2) establishing, also contrary to the view of many authorities, that the amygdala does not function in general stimulus-reward associations or primarily in negative emotions, but instead acts to update the current assessment of value based on drive states of both a positive and negative nature; and (3) showing that certain parts of the frontal lobe do not always need to be functioning for the inhibition of automatic behaviors to occur normally. These advances are briefly elaborated below, in turn. (1) The perirhinal cortex, a region located at the ventromedial aspect of the MTL, lies at the interface of the so-called ?MTL memory system? and the ventral visual stream. It receives inputs from both unimodal sensory areas such as area TE (which plays a key role in color and form vision) and the caudal, granular insular cortex (which plays a key role in touch), as well as from multimodal regions including the cingulate cortex and orbital frontal cortex. In studies of the perirhinal cortex, we found, contrary to the prevailing view of MTL function, that damage to each part of the MTL causes a unique set of behavioral deficits, some involving memory, others involving perception, and yet others involving response selection. The prevailing view of medial temporal lobe MTL function has two principal elements: first, that the MTL subserves memory but not perception, and second, that the many anatomically distinctive parts of the MTL function together in the service of declarative memory. Recent neuropsychological studies have, however, overturned both opinions. First, the perirhinal cortex represents information about objects for both mnemonic and perceptual purposes, including the conjunctions of features that compose individual objects, as well as the conjunctions of objects that compose visual scenes. Second, the idea that MTL components such as the hippocampus and perirhinal cortex contribute roughly equally to declarative memory has also been contradicted. The perirhinal cortex, but not the hippocampus, is essential for object recognition memory, the key test of declarative memory. By contrast, the hippocampus appears to be essential for processing information about places and paths, but it plays a small role, if any, in object recognition. These findings have been published in a follow-up of our original findings (Bussey, Saksida and Murray, 2006) and in two papers by Saksida, Bussey, Buckmaster and Murray in 2006, one in the journal Cerebral Cortex and the other in the journal Hippocampus. (2) Also in the past year, we studied the role of the amygdala and orbitofrontal cortex in affect and learning. The amygdala is widely accepted to be important for the recognition of negative emotions, such as fear, as well as for associating stimuli with aversive sensory inputs. Accordingly, research on the neuropsychology of the amygdala has focused largely on its role in learning about negative events, as assessed through fear conditioning paradigms. Its role in positive reinforcement has been relatively neglected, and sometimes denied. In contrast to the view that the amygdala functions primarily in negative reinforcement, our evidence, and those of others, shows that the amygdala also plays a central role in positive reinforcement. Accordingly, its contribution to positive emotions probably differs little, if at all, from its role in negative emotions. The amygdala endows the products of brain function with emotional valence by linking initially neutral neural representations with innate responses and performance rules. Higher brain functions so endowed include concepts and categories, learned response rules and strategies, analogies and inferences, and emotionally laden language. In addition, the amygdala functions in close relationship with the orbitofrontal cortex. The behavioral effects of amygdala lesions generally parallel those of orbitofrontal cortex lesions, and several investigators have emphasized the consistent association between the emotional changes and deficits in reversal learning and extinction that accompany either amygdala or orbitofrontal cortex damage. Dysfunctional interactions of the amygdala and orbitofrontal cortex with other regions are thought to underlie a host of psychiatric disorders, including clinical depression, bipolar disorder, obsessive-compulsive disorder, and schizophrenia. Our work has examined the specific and joint contributions of these two key components of the limbic system in order to correct misconceptions inherent in current medical thinking about their functions and to guide future research. We found that unilateral lesions of the amygdala-orbitofrontal cortex circuit disrupt affective processing, that the orbitofrontal cortex is critical for response selection based on predicted reward outcomes, regardless of whether the value of the outcome is predicted by affective signals (reinforcer devaluation) or by visual signals conveying reward contingency (object-reversal learning), that amygdala lesions facilitated the extinction of instrumental responses, and that lesions of the orbitofrontal cortex had the opposite effect (Izquierdo and Murray, 2006; Murray, Izquierdo and Malkova, 2006, chapter in preparation; Murray and Baxter, 2006). (3) In addition, we followed up an earlier study which found that subjects can master the reversed-contingency task, a task in which the smaller of two payoffs must be chosen in order to receive the larger of the two. We discovered that orbitofrontal lesions do not affect this aspect of inhibitory control (Chadusama, Kralik and Murray, 2006). In addition to these three main lines of research, we found that mental predictions, called prospective memories, are necessary in order to learn how to learn, a process called a ?learning set? (Murray and Gaffan, 2006).